COAST Risk Reduction Activities

1
COAST Risk Reduction Activities

• Curtiss O. Davis
• College of Oceanic and Atmospheric Sciences
• Oregon State University, Corvallis, Oregon 97331
• cdavis_at_coas.oregonstate.edu

2
HES-CW Products and Applications

• Products
• Spectral water leaving radiances
• Chlorophyll
• Chlorophyll fluorescence
• Turbidity
• Spectral absorption and scattering
• Applications
• Water quality monitoring
• Coastal hazard assessment
• Navigation safety
• Human and ecosystem health awareness (Harmful
  Algal Blooms)
• Natural resource management in coastal and
  estuarine areas
• Climate variability prediction (e.g., role of the
  coastal ocean in the carbon cycle)
• Nowcast and Forecast models of the coastal ocean

3
COAST and Risk Reduction Activities

• The Coastal Ocean Applications and Science Team
  (COAST) was created in August 2004 to support
  NOAA to develop coastal ocean applications for
  HES-CW
• Mark Abbott, Dean of the College of Oceanic and
  Atmospheric Sciences (COAS) at Oregon State
  University is the COAST team leader,
• COAST activities are managed through the
  Cooperative Institute for Oceanographic Satellite
  Studies (CIOSS) a part of COAS, Ted Strub,
  Director
• Curtiss Davis, Senior Research Professor at
  COAS, is the Executive Director of COAST.
• Initial activity to evaluate HES-CW requirements
  and suggest improvements
• Paul Menzel Presented GOES-R Risk Reduction
  Program at the first COAST meeting in September
  2004 and invited COAST to participate.
• Curt Davis and Mark Abbott presented proposed
  activities in Feb. 2005.
• CIOSS/COAST invited to become part of GOES-R Risk
  Reduction Activity beginning in FY 2006.
• Proposal Submitted to NOAA Sept 6, 2005.
• Activities to begin in May, 2006.

4
Risk Reduction ActivitiesPrincipal Roles of
Co-Investigators

• Curtiss Davis, program management, calibration,
  atmospheric correction
• Mark Abbott, COAST Team Leader
• Ricardo Letelier, phytoplankton productivity and
  chlorophyll fluorescence, data management
• Peter Strutton, coastal carbon cycle, Harmful
  Algal Blooms (HABs)
• Ted Strub, CIOSS Director, coastal dynamics,
  links to IOOS
• COAST Participants
• Bob Arnone, NRL, optical products, calibration,
  atmospheric correction, data management
• Paul Bissett, FERI, optical products, data
  management
• Heidi Dierssen, U. Conn., benthic productivity
• Raphael Kudela, UCSC, HABs, IOOS
• Steve Lohrenz, USM, suspended sediments, HABs
• Oscar Schofield, Rutgers U., product validation,
  IOOS, coastal models
• Heidi Sosik, WHOI, productivity and optics
• Ken Voss, U. Miami, calibration, atmospheric
  correction, optics
• NOAA/STAR
• Menghua Wang, atmospheric correction
• Mike Ondrusek, calibration, MOBY

5
HES-CW Data flow and Risk Reduction Activities
6
Approach to Algorithm Development

• Directly involve the ocean color community which
  has extensive algorithm development experience
  with SeaWiFS and MODIS
• NASA funded science teams developed, tested and
  validated calibration, atmospheric correction and
  product algorithms
• Additional product development and testing funded
  by U. S. Navy
• Documented in Technical reports, publications and
  Algorithm Theoretical Basis Documents (ATBDs)
• Algorithms are continuously evaluated and
  updated SeaWiFS and MODIS data routinely
  reprocessed to provide Climate Data Records with
  latest algorithms.
• Design program to assure compatibility of HES-CW
  products with VIIRS
• VIIRS algorithms based on MODIS ATBDs
• Similar calibration and atmospheric correction
  approaches
• Use the same ocean calibration sites for
  vicarious calibration
• Initial plans and algorithms based on SeaWiFS and
  MODIS experience modified to fit HES-CW in
  geostationary orbit.
• Advanced algorithms tested and implemented when
  available.
• Early tests planned using airborne hyperspectral
  data.

7
Risk Reduction Activities

• Approach to Algorithm Development
• Experience with SeaWiFS and MODIS
• Field Experiments to collect prototype HES-CW
  data
• Planned Risk Reduction activities
• Calibration and vicarious calibration
• Atmospheric correction
• Optical properties
• Phytoplankton chlorophyll, chlorophyll
  fluorescence and productivity
• Benthic productivity
• Coastal carbon budget
• Harmful algal blooms
• Data access and visualization
• Education and public outreach

8
Risk Reduction Plans Calibration

• Develop plan for on-orbit calibration
• At sensor radiance calibration must be /- 0.3
  to meet proposed Chlorophyll product accuracy
  requirement of /- 30
• Follow SeaWiFS and MODIS approach using moon
  imaging, solar diffuser and vicarious calibration
  to achieve this accuracy
• Risk reduction activity includes planning for
  highly accurate water leaving radiance
  measurements at two clear water ocean sites (NOAA
  ORA led effort)
• Additional coastal sites for validation of
  atmospheric correction in coastal waters and
  validation of coastal products
• Coordinated effort between NOAA ORA, CICS, CIOSS
• Good on-orbit calibration is only possible if the
  instrument is properly designed to provide stable
  accurate radiances over its lifetime. This must
  be demonstrated with good pre-launch calibration
  and characterization for MTF, stray light, etc.
• Support NOAA/NASA to provide feed back to
  instrument builders.
• Pre-launch calibration requirement is /- 5
  absolute, /- 0.5 channel to channel. The needed
  higher accuracies on-orbit can only be achieved
  with vicarious calibration.

9
Atmospheric Correction Challenges

• We anticipate three major challenges in
  developing the atmospheric correction algorithms
  for HES-CW.
• 1. Adaptation of the current algorithms for
  SeaWiFS and MODIS to the geostationary viewing
  geometry.
• 2. Dealing with Absorbing Aerosols which are
  common downwind from urban and industrial areas.
• 3. In coastal waters with high levels of
  suspended sediments, or large phytoplankton
  blooms the contributions at the NIR bands are not
  negligible. This can lead to significant
  underestimation of the satellite-derived
  water-leaving radiance spectrum (SeaWiFS, MODIS).

10
Current Atmospheric Correction Algorithms
SeaWiFS and MODIS algorithm (Gordon and Wang 1994)

• rw is the desired quantity in ocean color remote
  sensing.
• Trg is the sun glint contributionavoided/masked
  and residual contamination is corrected.
• trwc is the whitecap reflectancecomputed from
  wind speed.
• rr is the scattering from moleculescomputed
  using the Rayleigh lookup tables.
• ?A ra rra is the aerosol and Rayleigh-aerosol
  contributions estimated using aerosol models.
• For Case-1 waters in the open ocean, rw is
  usually negligible at 765 865 nm. ?A can be
  estimated using these two NIR bands.

Menghua Wang, NOAA/NESDIS/ORA
11
HES-CW Channels and Atmospheric Transmission
Windows
UV channels can be used for detecting the
absorbing aerosol cases Two long NIR channels
(1000 1240 nm) are useful for of the Case-2
waters
Menghua Wang, NOAA/NESDIS/ORA
12
Risk Reduction Plans Atmospheric correction

• Atmospheric correction needed to produce
  water-leaving radiance.
• Approach
• Evolution of algorithms from the current SeaWiFS,
  MODIS algorithms.
• Adjustments for Geostationary orbit geometry
• Adaptation to different spectral channels
• Development of coastal atmospheric correction
  algorithm
• Address absorbing aerosols,
• Address high reflectance in coastal waters where
  NIR channels cannot be used for aerosol
  calculations.
• Current effort between NOAA ORA, CICS, CIOSS
• Providing feedback to NOAA/NASA and instrument
  and spacecraft vendors to assure spectral channel
  characteristics, etc.
• Would like to expand effort to include
  collaborative efforts with CIMSS, CIRA, others?
• Explore advantages of using HES sounder and ABI
  data to improve atmospheric correction.

13
Risk Reduction Plans In-water Optical Properties
1

• Remote-sensing reflectance (Rrs, water-leaving
  radiance normalized by the downwelling
  irradiance) is a function of properties of the
  water column and the bottom,
• Rrs(?) fa(?), bb(?), ?(?), H, (1)
• where a(?) is the absorption coefficient, bb(?)
  is the backscattering coefficient, ?(?) is the
  bottom albedo, H is the bottom depth. In
  optically deep waters (when the bottom is not
  imaged),
• Rrs(?) fbb(?)/a(?) bb(?) (2)
• Where f is a proportionality constant that varies
  slightly as a function of the shape of the volume
  scattering function and the angular distribution
  of the light field. The backscattering
  coefficient bb(?) is the sum of the
  backscattering from the phytoplankton, detritus,
  suspended sediments and the water itself. The
  absorption coefficient a(?) is the sum of the
  absorption by CDOM, phytoplankton, detritus,
  suspended sediments and the water itself.

14
Risk Reduction Plans In-water Optical Properties
2

• Algorithms for SeaWiFS and MODIS use spectral
  channel ratios to calculate specific products,
  such as suspended sediments, chlorophyll and
  CDOM.
• This approach does not work if the bottom is
  imaged (e.g. West Florida Shelf), or in the
  presence of high levels of suspended sediments
  (e.g. Mississippi River Plume)
• Excellent Radiative Transfer Models (e.g.
  HYDROLIGHT) are available to model the light
  field the challenge for remote sensing is to
  invert those models to go from remote sensing
  reflectance to estimates of the in-water
  constituents.
• Two approaches are demonstrated that solve the
  full problem and produce values for water column
  optical properties, bathymetry and bottom type.
• A predictor-corrector approach is used to invert
  a semi-analytical model
• A look-up table approach has been used to invert
  HYDROLIGHT.

15
Bathymetry, Bottom Type and Optical Properties
Example Approach Semi-Analytical Models

• Semi-analytical model developed to resolve the
  complex optical signature from shallow waters.
• Simultaneously produces bathymetry, bottom type,
  water optical properties.

a) Bottom type and b) bathymetry derived from an
AVIRIS image of Tampa Bay, FL using automated
processing of the hyperspectral data. Accurate
values were retrieved in spite of the fact that
water clarity varies greatly over the scene.
(Lee, et al., J. Geophys. Research, 106(C6),
11,639-11,651, 2001.)
16
Bathymetry, Bottom Type and Optical Properties
Example Approach Look-up Tables
Interpretation of hyperspectral remote-sensing
imagery via spectrum matching and look-up tables.
Mobley, C. D., et al., Applied Optics, In Press.
17
Risk Reduction Plans In-water Optical Properties
3

• Planned Risk Reduction Activities
• NASA and the Navy have a set of band ratio type
  algorithms to produce in-water optical properties
  from SeaWiFS and MODIS data.
• Initial approach will be to adapt those
  algorithms for use with HES-CW.
• Main Risk Reduction effort will be to develop
  comprehensive methods along the lines of the Lee
  et al. and Mobley, et al. approaches that have
  been demonstrate for airborne hyperspectral data.
• Will work in all conditions even when the bottom
  is imaged
• Algorithm work can be initiated immediately with
  existing data sets but the HES-CW demonstration
  data set will be essential for the full
  demonstration of the algorithms.
• Initiate effort in 2006 to use existing data sets
  and to participate in the planning of the HES-CW
  demonstration experiment to assure that all of
  the essential data is collected.
• Expanded effort in 2009 utilizing the
  demonstration data set and Web based data system.

18
Risk Reduction Plans Phytoplankton chlorophyll,
chlorophyll fluorescence and productivity

• Chlorophyll and Chlorophyll fluorescence
• Fluorescence unambiguously associated with
  chlorophyll
• Signal is small, but use of baseline approach
  greatly reduces impact of atmosphere on
  retrievals
• Amount of fluorescence per unit chlorophyll
  varies as function of light, phytoplankton
  physiology, and species composition
• Validation relies on long time series of high
  quality measurements to ensure consistency
• IOOS, MOBY sites
• Analysis of MODIS Aqua and Terra data sets
• AVIRIS or other overflights
• Estimates of chlorophyll and productivity
• Continued field and satellite data analysis
• Modeling of quantum yield of fluorescence based
  on laboratory analyses, comparison with field
  measurements
• Incorporate quantum yield into productivity
  models
• Compare with recent chlorophyll/backscatter
  models using SeaWiFS

19
MODIS FLH bands avoid oxygen absorbance at 687
nm
20
MODIS Terra FLH vs Oregon optical drifters
derived FLH
21
Testing the MODIS FLH Algorithm
FLH vs. chlorophyll
FLH vs. CDOM
From Hoge et al.
22
Frequent measurements in morning can elucidate
quantum yield of fluorescence
23
Risk Reduction Plans Phytoplankton chlorophyll,
chlorophyll fluorescence and productivity

• Proposed activities
• - Development of chlorophyll and fluorescence
  algorithms based on SeaWiFS and MODIS legacy and
  modified to fit HES-CW in geostationary orbit.
• - Characterization of chlorophyll and chlorophyll
  fluorescence algorithm sensitivity based on
  HES-CW (waveband position and SNR)
  characteristics (i.e. Letelier and Abbott 1996)
• - Generation of HES-CW synthetic chlorophyll and
  fluorescence products in coastal (case II) waters
  using Hyperion and PHILLS data, and data from
  field experiments in 2007-2008.
• - These field experiments will serve to
• 1) Validate a chlorophyll algorithm for case II
  waters based on chlorophyll fluorescence.
• 2) Assess diurnal changes in algal physiology
  affecting carbonchlorophyll ratio and the
  chlorophyll fluorescence efficiency.
• 3) Evaluate how water column stability and
  CDOM concentrations affect the apparent
  relationship between chlorophyll concentration
  and the chlorophyll fluorescence in algorithms
  inherited from SeaWiFS and MODIS.
• 4) Develop improved productivity models
  incorporating laboratory estimates of quantum
  yield.

24
Risk Reduction Plans Harmful Algal Blooms
Background

• In the Gulf of Mexico, blooms of the toxic algae
  Karenia brevis result in shellfish bed closures
  and lost tourism that cost the state of Florida
  millions of dollars each year.
• Similar problems in other parts of the country
  with other toxic species.
• Ship based monitoring very expensive and time
  consuming
• Inadequate data frequently leads to unnecessary
  closings.
• HABSOS system being developed to provide early
  warnings using SeaWiFS data and models
• HES-CW will greatly improve warning systems like
  HABSOS
• More frequent data for cloud clearing
• Higher spatial resolution to assess conditions
  closer to the shell fish beds and beaches

25
Frequent sampling can assist in detection and
classification of HABs
Some properties have a diel cycle associated with
it. Documenting the diel dynamics can thus
potentially assist in documenting and identifying
material in the ocean
Case example
Detection of K. brevis
26
When K. brevis Blooms, conditions tend to be
calm. Under these Conditions the cells exhibit
a dramatic diel migration. The net result is a
10X increase in cells at the air-sea interface
over a several hour period. This unique feature
will be readily detected in HES-CW data.
27
HABSOS can immediately utilize improved spatial
resolution and frequency of coverage from HES-CW
28
Risk Reduction Plans Harmful Algal Blooms

• Proposed Risk Reduction Activities
• Improve methods for early detection of HABs from
  optical remote sensing data
• Not all HABs have a unique optical signature
  use additional information, e.g. vertical
  migration to identify blooms.
• Specific methods needed for each region of the
  country to identify local species, etc.
• Continue development of models of HAB dynamics
• Higher frequency of HES-CW data critical for
  cloud clearing and to include vertical migration
  in the models
• Prepare to use HES-CW data in warning systems,
  such as, HABSOS
• Increased frequency of sampling for cloud
  clearing will provide faster updates allowing
  more precise system for warnings
• Avoid unnecessary costly beach and shellfish bed
  closures
• Strong education component to educate the state
  and local managers and the public as to the
  benefits of HES-CW data and improved models and
  forecasts.

29
Coastal Carbon Cycle

• Detailed studies of the Oregon coastal upwelling
  system to determine its role as a CO2 source or
  sink.
• pCO2 in coastal (and other) environments is
  associated with characteristic chlorophyll and
  SST signatures.
• Using multiple satellite products and techniques,
  such as multiple linear regression, we have
  developed an approach to determine sea surface
  pCO2 from space.
• Combine this with winds from either
  scatterometer(s) or coastal/buoy meteorological
  stations to facilitate flux calculations.
• (Hales et al., 2004. Atmospheric CO2 uptake by a
  coastal upwelling system. Global Biogeochemical
  Cycles, 19, GB1009, 10.1029/2004GB002295.)

30
Coastal Oregon Study Site
31
Undersaturation of CO2 in coastal waters
Freshly upwelled water near the Oregon coast is a
CO2 source to the atmosphere. As the water moves
offshore the phytoplankton bloom making the same
waters a CO2 sink.
Cascade Head time series
32
Coastal CO2 Relationship to physics and biology
33
Risk Reduction Plans Coastal CO2 Fluxes

• The coastal ocean plays a large and poorly
  measured role in the global carbon cycle.
• Addresses NOAAs climate change goals
• HES-WC will provide valuable data to study this
  process
• Temporal sampling of 3 hours will enable basic
  budgets to be calculated and the tracking of
  processes such as productivity and subduction.
• This is a dynamic environment any ability to
  clear or alias clouds will enhance badly-needed
  coverage.
• Coupling with NASAs Orbiting Carbon Observatory
  (2008) will add significant coupling to
  atmospheric data.
• Proposed risk reduction activities
• Continue to develop and refine current models and
  algorithms using SeaWiFS, MODIS and shipboard
  data.
• Update algorithms to take advantage of HES-CW
  data.
• Adapt approach to take advantage of IOOS and
  associated modeling efforts.

34
Risk Reduction Plans Now-cast and forecast models

• Now-cast and forecast models are currently under
  development for the coastal ocean
• Model development will be closely coupled with
  IOOS,
• Current emphasis is on getting the physics right
  and on assimilating surface currents, wind data
  and other physical parameters,
• Some bio-optical models that could make excellent
  use of HES-CW data have been demonstrated,
• Work in this area will require the HES-CW
  demonstration data set to be collected in
  2007-2008,
• Plan to initiate modeling efforts in 2009.
• A second class of prognostic models for HABs are
  being developed for several coastal regions
• Begin limited effort in 2006 to support those
  models specifically emphasizing the utility of
  HES-CW data to improve skill of those models
• Utilize the HES-CW demonstration data set
  beginning in 2009.

35
EcoSim 2 Model Output for July 31, 2001 HyCODE
experiment at (LEO-15)
Satellite Measured
Bissett, et al., Submitted J. Geophys. Res.
36
Risk Reduction Plans Data Management

• Data processing, distribution and archiving
  issues.
• Need more processing capacity for atmospheric
  correction and product algorithms (3-5 X the
  calibration processing)
• Need for reprocessing with updated calibrations
  and new algorithms to make Climate Data Records
  and the need to archive CDRs
• Planned data system not sized for reprocessing.
• Next generation product generation and delivery
  services will build on the notion of web
  services, which are industry standard tools for
  building complex services from building block
  components and multiple data streams. Web
  services can provide new capabilities that are
  not anticipated in the original systems design.
  By designing these services as linked components
  rather than monolithic systems, GOES-R can
  provide a much greater degree of flexibility and
  evolution within a cost-constrained environment.
• We propose monitoring and providing advice on
  current plans for the HES-CW data system, with
  specific risk reduction activities beginning in
  2008.
• Web based server with the the Simulated HES-CW
  data from the proposed experiments. Include all
  ancillary data and access to models for testing.
  

37
BACKUP
38
Summary

• HES-CW will provide an excellent new tool for the
  characterization and management of the coastal
  ocean.
• Risk Reduction activities focus on calibration
  and algorithm development
• Initially provide SeaWiFS and MODIS heritage
  calibration and algorithms
• 2006-2008 field experiments to develop example
  HES-CW data for
• algorithm development and testing,
• Coordination with IOOS for in-situ data and
  coastal ocean models,
• Demonstrate terabyte web-based data system.
• Major focus on developing advanced algorithms
  that take advantage of HES-CW unique
  characteristics.
• Efforts coordinated with NOAA ORA, NMFS and NOS
  with a focus on meeting their operational needs.
• Special thanks to Ted Strub, Amy Vandehey and the
  COAST for their hard work getting this program
  started.
• Thanks to NOAA for funding and particularly to
  Stan Wilson, John Pereira, Eric Bayler and Paul
  Menzel for their support and guidance.

39
COAST proposed new HES-CW Channel Specifications
40
COAST proposed new HES-CW Channel Specifications
41
Frequency of Sampling and Prioritizing Goal
Requirements

• Threshold requirement is to sample all Hawaii and
  Continental U. S. coastal waters once every three
  hours during daylight
• Plus additional hourly sampling of selected areas
• Goal requirement is hourly sampling of all U.S.
  coastal waters is strongly recommended, for cloud
  clearing and to better resolve coastal ocean
  dynamics.
• Goal requirements compete with each other, e.g.
  higher spatial resolution makes it harder to
  increase sampling frequency or SNR.

• COAST top priority goals are
• Higher frequency of sampling
• Goal channels for atmospheric correction
• Hyperspectral instead of multispectral

HES-CW built to the threshold requirements will
be a dramatic improvement over present
capabilities for coastal imaging.
42
Proposed Experiments to collect simulated HES-CW
data (2 of 2)

• Experimental Design
• Choose sites with IOOS or other long term
  monitoring and modeling activities
• Intensive effort for 2 weeks to assure that all
  essential parameters are measured
• Supplement standard measurements at the site with
  shipboard or mooring measurements of
  water-leaving radiance, optical properties and
  products expected from HES-CW algorithms,
• Additional atmospheric measurements as needed to
  validate atmospheric correction parameters,
• As needed, enhance modeling efforts to include
  bio-optical models that will utilize HES-CW data.
• Aircraft overflights for at least four clear days
  and one partially cloudy day (to evaluate cloud
  clearing) during the two week period.
• High altitude to include 90 or more of the
  atmosphere
• 30 min repeat flight lines for up to 6 hours to
  provide a time series for models and to evaluate
  changes with time of day (illumination,
  phytoplankton physiology, etc.)
• All data to be processed and then distributed
  over the Web for all users to test and evaluate
  algorithms and models.

43
Chlorophyll and chlorophyll fluorescence of
optically-complex coastal waters

• MODIS Terra scene from 3 October 2001.
• The ratio of Fluorescence Line Height (FLH) to
  chlorophyll is a good indicator of the health
  of the phytoplankton population.
• FLH separates chlorophyll from suspended
  sediments in the Columbia River Plume.
• Fluorescence line height not available from
  VIIRS.

44
Monterey Bay Harmful Algal Bloom

• Harmful Algal Bloom in Monterey Bay threatens
  beach areas.
• Bloom near coast and on the order of 2 x 5
  km would not be resolved in 1 km VIIRS data.
• Additional channels on HES-CW aid bloom
  identification.

PHILLS-2 airborne hyperspectral data from Paul
Bissett, Florida Environmental Research
Institute. (October 2002 Ceratium spp. bloom)
45
HES-Will support the higher temporal and spatial
resolution required for coastal models
Satellite Measured
ECOSIM run for July 31, 2001 with ROMS Physical
model 15 minute time step and 300 m spatial
resolution (Paul Bissett, Florida Environmental
Research Institute)